基于单张图像的三维人脸建模与表情动画

在人机交互、数字娱乐等领域,传统的表情合成技术难以稳定地生成具有真实感的个性化人脸表情动画.为此,提出一种基于单张图像的三维人脸建模和表情动画系统.自动检测人脸在图像中的位置,自动定位人脸上的关键点,基于这些关键点和形变模型重建个性化三维人脸模型,对重建的人脸模型进行扩展得到完整的人脸网格,采用稀疏关键点控制的动画数据映射方法来驱动重建的人脸模型生成动态表情动画.实验结果表明,该方法稳定性强、自动化程度高,人脸模型与表情动画比较逼真.

3D face modeling and expression animation based on single image

In the area of human-computer interaction and digital entertainment,traditional techniques of facial expression synthesis can hardly produce realistic dynamic expression animation stably.A system of 3D face modeling and expression synthesis based on one face image is proposed.First,the face is detected in the image and the key points on the face are detected automatically.Then the individual 3D face model is reconstructed based on these key points and morphable models.Then the reconstructed face model is driven by animation data to produce dynamic expression animation based on sparse key points controlled animation data mapping method.Experimental results show that the method is robust and has a high degree of automation,the expression animation synthesized is realistic.